Client: (Client name changed for privacy.)
Role: Creative Director and Project Manager
Industry: EdTech / Language Learning
Located: USA/Japan
Year: June 2022 – June 2023
Team: Teachers, Writers, Illustrators, Voiceover Artists, Designers, Developers, and Business Analysts
Tools: ClickUp, React.js, Next.js
The founder built a TikTok audience of over a million followers by teaching Japanese in short clips. They understood what makes language learning feel approachable and that instinct for how people learn and what keeps them engaged was the asset around which the whole platform was developed.
In 2022, they needed someone to help turn that into something bigger. The audience was there, and demand was proven. Their question was how to convert a million casual followers into committed learners on a platform, and that’s where I came in.
Japanese is one of the harder languages to pick up. There are three writing systems, and the sentence structure works almost in reverse compared to English. For a beginner sitting down with a language app for the first time, the cognitive load alone is enough to make people quit immediately
The research we did early in the project kept surfacing the same three problems. Most tools relied heavily on memorisation, with almost no engagement to help learners through the difficult early stages. Many of the bigger platforms taught multiple languages, which diluted their ability to address the specific challenges Japanese presents. And without a learning community, beginners hit cognitive overload and drop off because they were learning alone with no sense of progress or momentum.
Meanwhile, their TikTok comments were a live feed of what the audience wanted. People who’d watched dozens of their videos were looking for something more structured, liked the approach, and wanted more of it. The community was the strongest indicator that the demand was genuine.
My role across the twelve months was Creative Director and Project Manager, which in practice meant sitting at the intersection of every discipline on the project. Teachers, writers, illustrators, voiceover artists, designers, developers, and business analysts. I kept all of them fixated on our shared vision, which was making sure the technical execution delivered what the creative direction intended.
The product decisions all came from the research. Three things came out of studying how the founder’s audience engaged with their content. We found that audio-based repetition dramatically improved recall; people who heard the language alongside reading it retained it faster. Visual associations also helped learners recognise Japanese characters more quickly than text-based approaches alone. And gamified progression, with visible momentum, rewards, and a sense of advancement, kept people coming back daily rather than dropping off after the first week.
The platform was designed as a module-based learning system. Learners sign in, select their proficiency level, and complete an initial icebreaker that captures their goals and motivations, which could be travel, work, relocation, or recreation. The platform uses that input to generate a personalised learning path rather than pushing everyone through the same sequence regardless of why they’re there.
Lessons combined audio with native Japanese and English voiceovers, flashcards for memory and pronunciation reinforcement, educational games that encouraged repetition, and visual illustrations to support character recognition. They also included offline worksheets for practice away from the screen. The LMS integrated all lessons, games, and assessments into a coherent progression that kept learners oriented and motivated.
On the technical side, the platform was built on React.js and Next.js, chosen for performance, options to scale and add features later, and the ability to deliver a responsive experience across devices. ClickUp was our coordination tool, and we used an Agile methodology for this project.
The founder was closely involved throughout. Their TikTok audience and Patreon community were a continuous feedback and survey channel for lesson length, pacing, format, which elements felt confusing, and which felt motivating.
A challenging part of the entire engagement that I didn’t fully anticipate at the beginning was sourcing Japanese UI/UX designers for a six- to eight-month contract.
Japanese voiceover artists were straightforward to find. Experienced UI/UX designers who matched the project’s creative vision and were willing to take on a short-term contract were not, and cultural factors played a significant role. Job loyalty and a strong preference for long-term employment meant the pool of designers available for this kind of engagement was limited. It required extensive outreach, flexible timelines, and workflow adaptation to bring the right people on and make it work.
Recognising that earlier would have meant more realistic expectations and a different recruitment strategy from the beginning. It was one of the clearest lessons of the project that some constraints are often human and not technical.
Months after the project began, an interactive language-learning platform with lessons, games, audio, illustrations, progress tracking, personalised learning paths, and a community structure was created around how our audience wanted to learn.
The foundation the founder had already created made this successful. A million followers is a form of product validation that most founders spend years trying to reach. They had this because they understood their learners, knew how to connect with them, and had built trust with the community.
